OVERVIEW

1
Isolating and Identifying Protein Interactions
within EGFR Complexes Laura Lee White1, 3,
Udayan Guha2, Harold E. Varmus2, Ileana M.
Cristea1 1Princeton University, Princeton, NJ
2Sloan Kettering NY, NY 3The Pingry School,
Martinsville, NJ
RESULTS
OVERVIEW
II. Quantitative Studies of EGFR Interactions
II. EGFR Quantification
Introduction Protein interactions are
responsible for the majority of cellular
processes. In this study, epidermal growth factor
receptor (EGFR) and associated proteins were
isolated to understand their involvement in lung
cancer. Aim To characterize signaling pathways
involved in lung cancer. Method Isolation of
protein complexes on magnetic beads, followed by
protein quantification and identification using
metabolic labeling and mass spectrometry. Results
We isolated wild type and mutant forms of EGFR
and identified its interacting proteins to gain
insights into their function.

• MS-Fit database search revealed EGFR as first
  hit.
• MOWSE score 2.33e10
• Protein Coverage 24.7
• MS-Fit searches heavy isotope labeled R and K.
• Isotopic labeling evident in database search.
• R and K highlighted modifications detailed in
  next figure.

I. Identification of wild type EGFR
Fig. 8
EGFR Sequences Wild type Mutated
Deletion L858R
E746 A750
Metabolic Labeling during
cell culture Combine Cell Lysates Isolate
EGFR Elution from Magnetic Beads
Fig. 4
Wild type, mutant (L858R), and deletion
(E746-A750) forms of EGFR were constructed to
study different interacting proteins. After
changing the genetic sequence of EGFR, the cells
were metabolically labeled. The process of
metabolic labeling incorporates stable
isotope-containing amino acids into newly
synthesized proteins. Growth medium is created in
which all light arginines and lysines were
replaced with heavy amino acids. The purpose
behind metabolic labeling is that each population
(wild type, mutation, deletion forms of EGFR, and
interacting proteins) are distinguishable through
MS analysis.
1
2
3
Mass spectra represents the distribution of
components (molecules or atoms) by mass-to-charge
ratio in a sample acquired by a mass spectrometer
100 kDa sample was cut from the SDS-PAGE into 1mm
pieces to digest the protein. Sample analyzed by
mass spectrometry. Monoisotopic peaks labeled in
spectra and control peptides eliminated.
Database search performed with all labeled
peptides. Search revealed EGFR peptides as most
prevalent in the sample. EGFR peptides are
marked in the shown spectrum.
Fig. 6
100 kDa Mass Spectra EGFR peptides labeled
kDa
deletion EGFR
Wild type EGFR
mutated EGFR
250
Heavy media
Light media
Heavy media
150
4 Da R 4 Da K
No modification
6 Da R 4 Da K
100
INTRODUCTION
75

• EGFR binds to epidermal growth factor and
  transforming growth factor a 1. It is a member
  of the ErbB family of receptors and, when
  activated, it signals mitogen-activated protein
  kinase signal transduction.
• Overexpression of EGFR is frequently seen in
  non-small cell lung cancer (NSCLC), the leading
  cause of fatal cancer in men and women.

Fig. 1
50
37
Fig 8. MS-Fit results. Highest hit, EGFR
protein. The modifications are indicated by the
database search engine.
25
15
Fig 1. EGFR image labeled with L858R mutation in
the activation loop and E746-A750 deletion in
exon 19. (J. Guillermo Paez, 2004).
Fig 6. Sample cut from SDS-PAGE, digested with
trypsin and analyzed by mass spectrometry EGFR
peptides labeled on the mass spectrum.
Fig. 2
Fig 4. Strategy for the isolation of mutated
EGFR and associated proteins via metabolic
labeling and magnetic bead isolation
Fig. 7

• Some patients with NSCLC have mutations in the
  EGF receptor that can be either short, in-frame
  deletions, or substitutions around the EGF
  tyrosine kinase domain.
• These patients have shown good response to
  gefitinib (Iressa), a drug that inhibits the
  receptor tyrosine kinase.
• In this study we examined the L858R mutation in
  the activation loop and E746-A750 deletion
  sequence in exon 19 (Fig. 1).
• NSCLC mutations are most frequent in Japanese
  patients, patients that are non-smokers, or
  those that have adenocarcinoma.

• X-Proteo Database Search revealed EGFR as top
  result
• High decision score 8.0
• EGFR peptides had the highest result based on the
  high decision score, the high number of assigned
  peptides, and an overall error reading consistent
  with the ToF calibration.
• Only wild type EGFR peptides are identified. No
  modifications due to metabolic labeling were
  included as search parameters when using the
  XProteo search engine.
• In order to relatively quantify the protein
  interactions with mutant EGFR, another database
  was used.

III. Protein Identification by Mass Spectrometry
Fig. 5
Protein Identification
Isolated proteins
Fig 2. Images of adenocarcinoma of the lung. Top
image is the control image of the tumor cell.
Bottom image is the tumor treated with gefitinib.
This treatment led to the death of most tumor
cells (image from Weisenthal Cancer Group, 2006
2).
Fig 9. Examples of identified EGFR peptides.
Labels show the isotopically modified R- and
K-containing peptides.

• Labeled EGFR peptides identified through MS-Fit
  search.
• Triple peaks verify EGFR metabolic labeling and
  the successful isolation of all the three forms
  of EGFR.
• Wild type, mutation, and deletion EGFR peptides
  are evident in sample.
• First peptides show the light isotopic peptide
  from wild type EGFR. Second and third peptides
  show heavy R and K amino acids within the
  mutated EGFR peptides.

METHODS
I. Approach for Isolating Protein Complexes
Fig 5. Protein Identification using Mass
Spectrometry. Example shown highlights the use of
a MALDI Q-ToF mass spectrometer for protein
analysis. Parts of the figure adapted from Steen
and Mann, 2004 4.
This approach proved successful for the isolation
and quantification of the three forms of EFGR.
We identified several interacting partners,
such as GRB2, and are in the process of
determining their relative levels of association
with the wild type, L858R mutated, and E746-A750
deletion forms of EGFR. Knowledge of proteins
that form complexes with the wild type and mutant
EGFR may lead to new therapeutic targets for
NSCLC.

• EGFR and associated proteins were isolated on
  magnetic beads conjugated with anti-EGFR
  antibodies (Fig. 3).
• Once isolated, the proteins were eluted and
  resolved by SDS-PAGE.
• Protein bands were cut from the gel and digested
  with trypsin.
• The peptides were analyzed by mass spectrometry.
• Mass spectra were labeled and the data was used
  to search databases to confirm EGFR and identify
  the interacting partners.

• Mass spectrometry identifies the chemical
  composition of a compound or sample on the basis
  of the mass-to-charge ratio (m/z).
• A MALDI quadrupole time-of-flight (Q-ToF) mass
  spectrometer was used to analyze peptides from
  proteins isolated with EGFR. The peptides are
  mixed with a matrix that absorbs at the
  wavelength of the laser and placed on a target.
  When dry, it creates a heterogenous distribution
  of matrix and peptide crystals which are then
  ionized.
• The ions travel through the quadrupole and enter
  the time-of-flight mass analyzer.

Acknowledgements
Luke De Bonnie Kaiser The Pingry
School Science Department Anna Arnaudo
Chase Palisch Val Carabetta
The Members of the Cristea Lab
References

1. Udayan Guha, Raghothama Chaerkady, Arivusudar
   Marimuthu, A. Scott Patterson, Manoj K. Kashyap,
   H. C. Harsha, Mitsuo Sato, Joel S. Bader, Alex E.
   Lash, John D. Minna, Akhilesh Pandey. Comparisons
   of tyrosine phosphorylated proteins in cells
   expressing lung cancer-specific alleles of EGFR
   and KRAS. (2008)PMAS, in print.
2. Larry Wiesenthal, M.D., P.h.D., Wiesenthal Cancer
   Group.(2006). Test identifies patients who
   benefit from targeted cancer drugs. Europeoan
   Hospital Journal.
3. Ileana M. Cristea, John-William N. Carroll,
   Michael P. Rout, Charles M. Rice, Bran T. Chait,
   and Margaret R. MacDonald., (2001). Tracking and
   Elucidating Alphavirus-Host Protein interactions.
   Joournal of Biological Chemistry. Volume 281, 40.
   30269-30278
4. Hanno Steen, Matthias Mann.. The ABCS (and
   XYZS) of Peptide Sequencing. (2004) Nature
   Reviews. September. Vol 5, 699-711

Fig 7. Database search of known masses from
labeled mass spectra. EGFR is highest match of
peptides in sample.
Fig 3. Approach for studying protein
interactions 3